I have the impression that there are several problems together.
May be removing the cleanUpforRelease and see if it fixes the problems with
dateandtime delays is a good move.
Stef
On Feb 27, 2013, at 1:57 PM, Camillo Bruni camillobr...@gmail.com wrote:
To rule out side-effects, I suggest
Hmm…
https://ci.inria.fr/pharo/job/Pharo-2.0-Tests/label=linux/187/consoleFull
Smalltalk stack dump:
0xbfb19878 M WideString(Object)becomeForward: 0x795d44e4: a(n) WideString
0xbfb19894 M WideString? 0x795d44e4: a(n) WideString
0xbfb198b4 M WriteStreamnextPut: 0x795c5f8c: a(n) WriteStream
I will setup a second Pharo-2.0-Test running on the StackVM so we can compare
them bit.
On 2013-02-27, at 13:40, Marcus Denker marcus.den...@inria.fr wrote:
Hmm…
https://ci.inria.fr/pharo/job/Pharo-2.0-Tests/label=linux/187/consoleFull
Smalltalk stack dump:
0xbfb19878 M
yes, i seen that dump, Marcus.
It is of course easy to blame become, but it can be worse:
something corrupts heap, and the corruption stays dormant, until
become operation, because
it scans heap and then stumbling upon invalid pointer, which leads to crash.
On 27 February 2013 13:47, Camillo
To rule out side-effects, I suggest deactivating ALL NB tests and see if we
still have such unpredictable behavior. Since last november we basically didn't
manage to get the system back into stable mode.
And there are currently three things (maybe unrelated) that worry me a lot:
- delay
On 27 February 2013 13:57, Camillo Bruni camillobr...@gmail.com wrote:
To rule out side-effects, I suggest deactivating ALL NB tests and see if we
still have such unpredictable behavior. Since last november we basically
didn't
manage to get the system back into stable mode.
why not. won't